Possible bug in VectorIndexRetriever?

I've tried both llamaindex 0.6.7 and 0.7.1 -- but in both the node_ids and doc_ids filters are not working at all for this query. No matter what, it always returns all nodes in the index, it doesn't filter them only to the node_ids or doc_ids provided in the parameter, which is how the documentation suggests to use it.

VectorIndexRetriever(index, similarity_top_k=10,doc_ids=list_of_doc_ids).retrieve("hello")
VectorIndexRetriever(index, similarity_top_k=10,node_ids=list_of_node_ids).retrieve("hello")

It's baffling. It's a complete blocker, going to turn to Langchain indexes only to try and solve.
Even if I wait, when I try again I still hit the limit, even for a very small number of document that are each very small

Hey there...
I'm trying to use the CSV Pandas dataloader, but am stuck on this error, apparently in the chain_wrapper.py script. Any chance on a fix for this anytime soon? or a workaround? https://github.com/jerryjliu/gpt_index/issues/601

from llama_index import download_loader

---> 7 import openai
8 from langchain import Cohere, LLMChain, OpenAI
----> 9 from langchain.llms import AI21

ImportError: cannot import name 'AI21' from 'langchain.llms'

I've read through all the docs, but I'm having trouble translating the concepts you describe in terms of the use-cases in general. I'm a datascientist with a good amount of NLP experience but still having some trouble mapping the different concepts into GPT Index.
If the source data isn't essay data, will I still benefit a good amount?